timeouts on non-virtual builders lead to cascade failure across the build farm
Bug #419317 reported by
LaMont Jones
This bug affects 1 person
| Affects | Status | Importance | Assigned to | Milestone | |
|---|---|---|---|---|---|
| Launchpad itself |
Triaged
|
Low
|
Unassigned | ||
Bug Description
The most common case for a non-virtual builder timing out during a build is because the build used sufficient machine resources to starve communication with the build-master.
Therefore, it is better to require human interaction to retry the build, than to automatically rescue the build and try it on the next builder in line, only to take that one out of commission as well.
Obviously, virtual builders, since they can go away at any time, should continue to have rescue efforts on behalf of the builds that are affected by hardware going away.
thanks,
lamont
| Changed in soyuz: | |
| status: | New → Triaged |
| importance: | Undecided → Medium |
| tags: | added: soyuz-build |
| Changed in launchpad: | |
| importance: | Medium → High |
| summary: |
- non-virtual builders that timeout should not result in rescued builds + timeouts on non-virtual builders lead to cascade failure across the + build farm |
Revision history for this message
| Julian Edwards (julian-edwards) wrote | #1 |
| Changed in launchpad: | |
| importance: | High → Wishlist |
| importance: | Wishlist → Low |
To post a comment you must log in.
This is a lot less of an issue now that we have failure counting. A single bad build won't take out the whole build farm.